linux-kernel - Re: [RFC V2] printk: add warning while drop partial text in msg

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <CAOVJa8H-8-Oh=1NZRj1EZYeVzmAeozHZDf2LSZ3qYX1O7HDUdw@mail.gmail.com>
Date:   Wed, 27 Sep 2017 21:59:52 +0800
From:   pierre kuo <vichy.kuo@...il.com>
To:     Sergey Senozhatsky <sergey.senozhatsky.work@...il.com>
Cc:     Petr Mladek <pmladek@...e.com>,
        Sergey Senozhatsky <sergey.senozhatsky@...il.com>,
        rostedt@...dmis.org, linux-kernel@...r.kernel.org
Subject: Re: [RFC V2] printk: add warning while drop partial text in msg

hi:
> printk_deferred("%s", local_string[2048]) makes no sense anyway,
> since we limit the message size to 1024 - HEADER chars in
> vprintk_emit()  // see static char textbuf[LOG_LINE_MAX].

In local_string[2048], it will all be 0 from [590] to [2047].
And vprintk_emit() will cut message from [LOG_LINE_MAX(992)] to [2047].

so, from [0] to [589], whole messages are saved in msg->text as
expected, from "this is the 0 line\n" to "this is the 29 line\n"
No truncating happen in this stage.
(sorry for declaring local_string[] as 2048 makes people confused.)

Below is what msg->text saved:
560 + 30 = 590
^^      ^^^
 |        |
 |        --> 30 "\n"
 |
 --> 560 chars without "\n" from  "this is the 0 line" to "this is the 29 line"

Below is complete message size output from msg_print_text:
18*30 + 590 = 540 + 590 = 1130 Bytes
^^^^^     ^^^
  |           |
  |           --> 590 chars of msg->text
  |
  -->30 lines with each 18 chars in the prefix(15 for timesatmap and 3
for levle and facility)

Below is the constituents that msg_print_text sent to buf[1024]
(the size of buf, 1024, pass to msg_print_text is < 1130)
18*27 + 530= 486 + 530=  1016 Bytes
^^^^^     ^^^
  |           |
  |           -->max chars found by memchr with "\n" as copying unit
to meet the upper bound, 1024.
  |               (from "this is the 0 line" to "this is the 26 line")
  |
  --> 27 lines with each 18 character in the prefix

so characters left in msg->text not putting to the buffer are 590 -
530 = 60 chars.
(Truncate the message from "this is the 27 line\n" to "this is the 29 line\n".)

We use the example in this mail since we try to collect the message at
different places in our driver.
And batch to printk for saving individual output time and group
message together.

> I'm not quite following what were you trying to prove, sorry.
>does any function in the upstream kernel printk()-s buffers
>larger than LOG_LINE_MAX? which one?

You are correct.
The upstream kernel printk()-s buffers are indeed smaller LOG_LINE_MAX
and not with multi "\n" like this example did.

> we are straggling to resolve the _existing_ printk issues, so
> _theoretical_ and never seen problems are not on my radar.

Got it and really appreciate your kind explanation.